Background Smoking has been implicated as a risk factor for the development of Rheumatoid Arthritis (RA), especially for the ACPA-positive subgroup. Whether smoking also affects the severity of RA is less clear as both positive and negative studies have been published.

Objectives We aimed to study the effect of smoking on the progression of joint destruction by evaluating several longitudinal cohorts.

Methods Smoking status was assessed in total in 2766 patients with on average 3 radiographs: 703 RA-patients (3,656 X-rays) included in the Leiden Early Arthritis clinic, in 839 patients (2,870 X-rays) included in the Swedish BARFOT-study, in 339 patients (1,062 X-rays) from Wichita (US) and in 885 patients (885 X-rays) from the National Databank of Rheumatic Diseases. In the Dutch and Swedish patients, data on present, past and never smoking were available. In the North American patients, presence versus absence of smoking was assessed. All X-rays were scored according to the Sharp-van der Heijde Score (SHS). The majority of patients were recruited in an era when early, aggressive treatment was still uncommon.

The association between smoking and the progression of joint destruction was analyzed using a multivariate normal regression analysis with the radiographic score as response variable for the patients with repeated measurements available and using a linear regression analysis for the patients with a single time-point measurement. All analyses were adjusted for age, gender and treatment. An inverse variance random effects combined analysis was performed using STATA.

Results When comparing radiological joint damage for RA-patients that were never, past and current smokers, smoking was associated with a more severe disease course in the Dutch (p=0.042) and Swedish (p=0.022) patients. In the two North American cohorts, no significant associations were found between smoking and joint damage. A combined analysis showed a significant effect of smoking on joint damage (p=0.021). With regards to the risk of development of RA, it is known that smoking in particular predisposes to the development of ACPA and that these antibodies are associated with RA. The subsequent question was to what extent ACPA mediates also the association between smoking and joint damage. Therefore the analyses were repeated with ACPA as additional adjustment factor. Then, no significant results were obtained in any of the cohorts.

Conclusions This study comprising four cohorts of in total 2766 RA-patients and 8473 X-rays showed that smoking is a risk factor for the progression of joint destruction, but not independent of ACPA-status.

Disclosure of Interest None Declared